Cancer classification and identification are major areas in medical research. DNA microarrays could provide useful information for cancer classification at the gene expression level. The number of genes in a microarra...
详细信息
Cancer classification and identification are major areas in medical research. DNA microarrays could provide useful information for cancer classification at the gene expression level. The number of genes in a microarray is always several thousands while the number of training samples always several dozens. In such case most of the machine learning models suffer from the overfitting and it is necessary to select a handful of most informative genes. An adaptive and iterative gene selection algorithm based on least squares support vector machines is proposed in this paper. The algorithm adopts sequential forward selection search scheme. The number of selected genes can be determined adaptively. The total number of genes processed by the proposed algorithm is smaller than that processed by other algorithms using support vector machines. Results of numerical experiments show that the proposed algorithm trains fast and achieves comparable performance on two well-known benchmark problems.
Moving object segmentation in the compressed domain plays an important role in many real-time applications, e.g. video indexing, video surveillance, etc. H.264/AVC is the up-to-date video coding standard, which employ...
详细信息
ISBN:
(纸本)300018726X
Moving object segmentation in the compressed domain plays an important role in many real-time applications, e.g. video indexing, video surveillance, etc. H.264/AVC is the up-to-date video coding standard, which employs several new coding tools and provides a different video format. In this paper, a robust spatio-temporal segmentation approach to extract moving objects in the compressed domain for H.264 video stream is proposed. Coarse moving objects are firstly extracted from the motion vector field through the MRF classification process. Spatial segmentation is based on the HT coefficient, which is used to refine the object boundary of temporal segmentation results. Fusion of spatio-temporal segmentation is based on region projection based labeling technique. Further edge refinement is performed to smooth the object boundaries. The performance of the proposed approach is illustrated by simulation carried on several test sequences. Experimental results show that the proposed approach can provide the visually similar results as by using spatial domain process with much less computational complexity.
To use role-based access control (RBAC) in wireless network is difficult than that in wired network. RBAC needs to search relative tables to get the user's permissions. We present an access control judgment algori...
详细信息
To use role-based access control (RBAC) in wireless network is difficult than that in wired network. RBAC needs to search relative tables to get the user's permissions. We present an access control judgment algorithm which bases on artificial neural network (ANN). The algorithm reduces the data transmission using bit string to express roles and permissions. The algorithm employs set theory to represent roles and their inheritance hierarchy, as well as conflicted permissions. It uses selected roles as input vectors and the matching permissions which contain no conflict as the output vectors to train the ANN. Then it uses the trained ANN to compute directly users' permissions when the system is under running condition, instead of searching tables. That improves the efficiency of access control. The algorithm is simple and efficient, which makes it easy to be realized in wireless networks
The self-organizing management mode in P2P networks leads to a large amount of selfish behaviors among peers. The corresponding solutions to this problem could hardly keep the merits of P2P network simultaneously, suc...
详细信息
The self-organizing management mode in P2P networks leads to a large amount of selfish behaviors among peers. The corresponding solutions to this problem could hardly keep the merits of P2P network simultaneously, such as self-organization or dynamic scalability. Most of the proposed reputation management models or incentive ones use the flooding mechanism to learn historical behaviour information of other peers, which causes excessive incurred packets, and thus limits the dynamic scalability. A novel incentive framework named as ResP2P based on random matching games theory is given in the paper. Peer reputation and its renewal mechanism, along with some essential social norms are introduced in ResP2P model, which stimulates rational peers to maximize their own utility and contribute their free resource. Experiments have verified the validity and efficiency of the incentive mechanism. The relative distributed algorithm can easily be deployed in a P2P networks and satisfied with self-organization and scalability.
Structured P2P networks create a virtual topology on top of the physical topology. The only relation between the two layers is the hashing algorithm, which makes the node's logical ID independent of its physical l...
详细信息
Structured P2P networks create a virtual topology on top of the physical topology. The only relation between the two layers is the hashing algorithm, which makes the node's logical ID independent of its physical location. By analyzing the Hash function, some novel logical connections among the destination node, the traditional semantic routing relay node sequence, and the ID of the clustering neighboring nodes are found. In this paper, the SCSRAA (self-organizing clustering semantic routing advanced algorithm) is resented to improve the efficiency of semantic routing. Since the clustering nodes only have local views in self-organizing mode, some rules are proposed for a node to learn other nodes' physical location. The SCSRAA's routing algorithm is described completely. Simulations have verified that the method can improve the semantic routing efficiently.
Tô et al. presented a bilinear-map-based traitor tracing scheme with revocation, but it is a symmetric scheme because it does not provide non-repudiation. An improved Tô scheme was proposed using oblivious p...
详细信息
Tô et al. presented a bilinear-map-based traitor tracing scheme with revocation, but it is a symmetric scheme because it does not provide non-repudiation. An improved Tô scheme was proposed using oblivious polynomial evaluation (OPE) protocol and service parameters. Under the precondition of general sameness capabilities of both Tô's and improved Tô's scheme, the new scheme adds some advantages such as providing multi-service capability, user's non-repudiation and data provider's no framing innocent users. Furthermore, it is also proved to be semantically secure under the decisional bilinear Diffie-Hellman problem (DBDH) assumption.
Semantic routing is one of key parts in P2P routing researches. The intelligent search mechanisms support flexible semantic expression but hold low scalability and recall rate. Contrastively, semantic overlay network ...
详细信息
Semantic routing is one of key parts in P2P routing researches. The intelligent search mechanisms support flexible semantic expression but hold low scalability and recall rate. Contrastively, semantic overlay network is scalable but either is difficult to organize or take too large maintenance spending. This paper proposes a new P2P semantic routing model in which node array is organized by match path and probability balance tree, and then an approximately balance distributed structure is obtained. All nodes will take routing decision according to query content while limiting maintenance spending to relative low level. The model supports flexible semantic search and high scalability, and ensures that each node can reach any corner of the network. The model runs with no center service in which all nodes simultaneously take index storage and data storage, and forward task to share system's running load by only maintaining a little local information.
We present a robust disaster recovery system model according to the requests of disaster recovery talent of a key e-government project of China. The model strengthens the safe communication ability and guarantees the ...
详细信息
Assuming a specific network service provider is in the position of natural monopolization, the pricing strategy most likely adopted by the provider is analyzed with the objective of profit maximization. The optimal pr...
详细信息
Assuming a specific network service provider is in the position of natural monopolization, the pricing strategy most likely adopted by the provider is analyzed with the objective of profit maximization. The optimal pricing problem is decomposed into two sub-problems regarding user demands and marginal cost respectively. The utility characteristics of transport control protocol are taken into account to infer a typical user's transport demands from the principle of net-utility maximization. Then the total demands function is deduced on the assumption that the user's budget constraint subjects to negative exponential distribution. With respect to marginal cost, the active queue management rationale is used for reference. Thereafter the relationship between marginal cost and network congestion cost is carefully examined, from which a piecewise marginal cost function in power form is presented. This pricing mechanism gives attention to the profit maximization objective of network service providers as well as the technical characteristics of transport control protocol. We believe that it would conduce to more efficient network resource allocation.
For group-oriented applications, designing secure and efficient group key management schemes is a major problem. We present a group key management scheme for dynamic peer networks, which supports join, leave, merge an...
详细信息
ISBN:
(纸本)1424406048
For group-oriented applications, designing secure and efficient group key management schemes is a major problem. We present a group key management scheme for dynamic peer networks, which supports join, leave, merge and partition events. In the scheme, the numbers of rounds and messages are close to the lower bounds of those for dynamic group key management, and the length of messages and computation costs are less than those of the existing schemes. Furthermore, this scheme provides forward secrecy, backward secrecy and key independence
暂无评论